Phil Young: Don’t base retirement advice on past performance

Phil-Young-700x450.jpg

Nobody likes probability. Nobody wants to talk about it. Yet it is unavoidable. The vast majority of people with financial advisers are 55 or older, which means managing income in retirement is their main priority. Probability is very important for them and, consequently, very important for advisers. There is little room for error.

The number one concern should be running out of money. We do not know when we will die but we can use averages. That said, we know averages are deeply flawed. Health information can help but it is only an indicator. We can control what we spend and roughly estimate what that might buy, assuming we take enough risk to beat inflation, but can we make reasonable predictions about future investment performance should that risk be taken?

Economic assumptions are pretty bleak right now, with uncertainty the predominant theme. What we can take from 2008 is that modern portfolio theory, which most advisers based their investment advice on, is flawed. It does not work in severely stressed conditions.

Behavioural economics has provided us with some good explanations as to why this happens but offers no comfort on the future. Theories and philosophies are created after the fact, as we attempt to explain how an event we had not thought possible happened. Unlike modern portfolio theory, behavioural economics offers no alternative investment model for investors. In this respect, it is a pessimistic theory. So diversification into non-correlating asset classes brings probability – not certainty – of success.

There is also ongoing academic debate around safe withdrawal rates. Four per cent has been challenged as being based on US data, and there are suggestions that 3 per cent or 3.5 per cent may be more advisable in the UK. Is this overly cautious because of current economic uncertainty or a genuine long-term view?

The problem of averages again raises its head. If 4 per cent is the correct figure, this will take into account a number of assumed economic cycles and a steady withdrawal rate regardless of market conditions. However, a client with a good amount of capital can keep two years’ worth of cash available to draw on during poor market conditions, which makes a higher rate of withdrawal safer, along with a higher level of equity exposure, allowing for a quicker recovery.

The somewhat counter-intuitive process of higher equity exposure in retirement seems to be the favoured approach borne out by academic research for those not wanting to lock into annuity rates. The question is, with no idea about the depth or longevity of future market dips or crashes, all we have is probability based on past experience.

The FCA was mocked for mentioning “rules of thumb” in its Financial Advice Market Review report but I am not sure we have much more, albeit there is some good research sitting behind them.

Stochastic tools have an edge over deterministic ones when looking at income in retirement projections, if only in that they lay the probabilities bare. As these are still based on past performance you need to consider the possibility that any future crash may be different from anything experienced before. In addition, stochastic tools tend to model future investment performance but the data needs to be taken back into a conventional deterministic cashflow model to take into consideration how other assets, such as cash on deposit or selling a property, could be used in the future for income or capital as well.

In short, nothing works perfectly. I have counted a dozen different probabilities mentioned in this article already, some of which are contingent on other risks. For anyone uncomfortable with this, you have the option of paying to protect against uncertainty, volatility or capital risk. This could be through an annuity purchase or a product with guarantees, for example.

An advice fee should ideally provide an alternative form of reassurance, not through formal guarantees (there will always be risk) but by constantly reviewing and optimising a client’s withdrawal strategies so they are “better than average”. This, and not investment performance, is a valid test of an adviser’s technical expertise and ability to manage client behaviour. Product based solutions remain two dimensional, focussing only on the element of a client’s portfolio they can control, such as blending an annuity with drawdown or offering guarantees, as they do not have the full range of tools and assets at their disposal. They are also expensive. We have seen little take up from advisers of “third way” contracts for that reason.

I do not see room for both the cost of fully automated solutions and the full cost of ongoing financial advice, as the numbers simply do not add up. More expensive, more secure products should be designed for DIY investors or transactional advice.

The priority for innovation for advisers should not be products but the tools and information which help to consider, model and explain the various probabilities each client faces when trying to ensure their money does not run out in retirement. I am pleased to say I have started to see some evidence of this beginning to happen.

Phil Young is managing director at Threesixty